Slide 1

(2) Dipartimento di Scienze dellInformazione, Universit di Bologna (!) Semantic Technology Laboratory ISTC-CNR Fine-tuning triplication with Semion Andrea Giovanni Nuzzolese (1) andrea.nuzzolese@istc.cnr.it Aldo Gangemi (1) aldo.gangemi@cnr.it Valentina Presutti (1) valentina.presutti@cnr.it Paolo Ciancarini (1,2) ciancarini@cs.unibo.it Lisbon, October 15 2010

Slide 2

Outline Motivations The transformation method The tool Conclusion and future work

Slide 3

The Web of Data is fed by triplifiers, tools able to transform content to Linked Data The Web of Data is fed by triplifiers, tools able to transform content to Linked Data Triplifiers implement various methods typically based on bulk recipes which allow for no or limited customization of the process Triplifiers implement various methods typically based on bulk recipes which allow for no or limited customization of the process Lack of good practices for knowledge representation and organization Lack of good practices for knowledge representation and organization The transformation relies on predetermined implicit assumptions on the domain semantics of the non-RDF data source The transformation relies on predetermined implicit assumptions on the domain semantics of the non-RDF data source Motivations

Slide 4

each table is a rdfs:Class each table record is an owl:Individual each table column is a rdf:Property A common recipe

Slide 5

limited customization of the transformation process (e.g. a user cannot map a table to a property) difficulty in adopting good practices of knowledge reengineering and ontology design (e.g. ontology design patterns each table column to a rdf:Property limited exploitation of OWL expressivity for describing the domain Implications

Slide 6

Some comparisons

Slide 7

Slide 8

A meta-model for RDBs

Slide 9

and for XML

Slide 10

Example of transformation: DB Reengineering Refactoring to LMM

Slide 11 Daily Maximum Tempera"> Daily Maximum Temperature 38 Daily Minimum Temperature 12 Hourly Probability of Precipitation 27 Weather Type, Coverage, and Intensity Example XML"> Daily Maximum Tempera" title=" Daily Maximum Tempera">

Daily Maximum Temperature 38 Daily Minimum Temperature 12 Hourly Probability of Precipitation 27 Weather Type, Coverage, and Intensity Example XML

Slide 12

Semion refactoring The refactoring allows to align a data set expressed with a specific vocabulary/ontology to another vocabulary/ontology is expressed as a set of rules expressed in SWRL rules realize recipes that can be saved (refactoring patterns)

Slide 13

The Linguistic Meta-Model (LMM) LMM plays the role of a mediator ontology LMM allows a semiotic cognitive representation of knowledge based on the so-called semiotic triangle LMM allows a semiotic cognitive representation of knowledge based on the so-called semiotic triangle most knowledge representation schemata can be aligned to the semiotic triangle most knowledge representation schemata can be aligned to the semiotic triangle

Slide 14

The reengineering perspective

Slide 15

The refactoring perspective

Slide 16

Conclusion and future work